Historics query returns different output every time


I’m calling the historics API through your python wrapper.
I’m interested in receiving the 100% of tweets mentioning a twitter id or retweeting a tweet from the same ID.
This is the csdl:

(twitter.mention_ids in [15155074] or // mentions
twitter.retweeted.user.id in [15155074] or // retweeted
twitter.user.id in [15155074]) // user tweets
AND interaction.sample <= 100
I built an historic object using 100 as sample parameter and with a 6 hours difference between end and start timestamps (in November).
I was expecting receiving exactly the same data if launching the same query again, as I’m asking for the 100% data for this specific query and I’m not changing anything.
Instead I receive every time a different amount of tweets (167 tweets the first time, 222 the second, 198 tweets the third).
How is that possible?
Usually I have a list of twitter_ids instead of the only one in this example.



This sounds like something best handled with a support ticket. If you head to our support site and raise a ticket providing the full details of the query (such as username, historic id, time stamps etc) we can look in to this further.